Estimated generic prices of cancermedicines deemed cost-ineffectivein England: a cost estimation analysis

Andrew Hill,1 Christopher Redd,2 Dzintars Gotham,3 Isabelle Erbacher,3

Jonathan Meldrum,4 Ryo Harada5

To cite: Hill A, Redd C,Gotham D, et al. Estimatedgeneric prices of cancermedicines deemed cost-ineffective in England: a costestimation analysis. BMJOpen 2017;7:e011965.doi:10.1136/bmjopen-2016-011965

▸ Prepublication history andadditional material isavailable. To view please visitthe journal (http://dx.doi.org/10.1136/bmjopen-2016-011965).

Received 30 March 2016Revised 4 October 2016Accepted 17 November 2016

1Department ofPharmacology andTherapeutics, University ofLiverpool, Liverpool, UK2Peninsula College ofMedicine and Dentistry,Peninsula Medical School,Universities of Exeter andPlymouth, Plymouth, UK3Faculty of Medicine, ImperialCollege London, London, UK4Faculty of Medical Sciences,University College London,London, UK5Department of Economics,University of Cambridge,Cambridge, UK

Correspondence toDzintars Gotham;dg1911@ic.ac.uk

ABSTRACTObjectives: The aim of this study was to estimatelowest possible treatment costs for four novel cancerdrugs, hypothesising that generic manufacturing couldsignificantly reduce treatment costs.Setting: This research was carried out in a non-clinical research setting using secondary data.Participants: There were no human participants inthe study. Four drugs were selected for the study:bortezomib, dasatinib, everolimus and gefitinib. Thesemedications were selected according to their clinicalimportance, novel pharmaceutical actions and theavailability of generic price data.Primary and secondary outcome measures:Target costs for treatment were to be generated foreach indication for each treatment. The primaryoutcome measure was the target cost according to aproduction cost calculation algorithm. The secondaryoutcome measure was the target cost as the lowestavailable generic price; this was necessary whereexport data were not available to generate an estimatefrom our cost calculation algorithm. Other outcomesincluded patent expiry dates and total eligible treatmentpopulations.Results: Target prices were £411 per cycle forbortezomib, £9 per month for dasatinib, £852 permonth for everolimus and £10 per month for gefitinib.Compared with current list prices in England, thesetarget prices would represent reductions of 74–99.6%.Patent expiry dates were bortezomib 2014–22,dasatinib 2020–26, everolimus 2019–25 and gefitinib2017. The total global eligible treatment population in1 year is 769 736.Conclusions: Our findings demonstrate thataffordable drug treatment costs are possible for novelcancer drugs, suggesting that new therapeutic optionscan be made available to patients and doctorsworldwide. Assessing treatment cost estimationsalongside cost-effectiveness evaluations is animportant area of future research.

INTRODUCTIONIn 2013, there were 8.3 million cancer deathsworldwide, representing 15% of overall mor-tality.1 There were an estimated 14 million

incident cases in 2012, a figure that isexpected to rise to almost 24 million by2035.2 Most diagnoses occur in low andmiddle-income countries (LMICs). In 2009,the worldwide cost of incident cancer casesalone was estimated to be $286 billion.3 Overthe past decade, several new classes of cancerdrugs have entered markets across theworld.4

The high prices of new cancer treatmentsare known to be a barrier to access inLMICs, where monthly drug prices oftenexceed annual incomes.5 These prices havebegun to pose problems in high-income set-tings too: newer drugs are a major contribu-tor to the 10-fold increase in the average costof cancer treatment in the UK since 1995.6

Drug prices account for roughly a quarter ofall cancer costs and prices have increased 10times in the past decade.7 Price is a keyfactor behind disparities in cancer healthcarein Europe, where €13.6 billion was spent oncancer drugs in 2009, amounting to 27% ofall cancer care costs.8 9

While cancer medication costs continue torise, there is only a weak correlation with

Strengths and limitations of this study

▪ A conservative and inefficient manufacturingmodel was used to generate realistic targetprices. Generic prices represent real-world marketcosts, which are likely to decrease in the future.

▪ We used peer-reviewed, publicly available epi-demiological data to generate robust eligibletreatment populations.

▪ The estimated treatment costs assume theabsence of intellectual property monopolieswhich, for drugs under patent protection, maynot be possible for several years.

▪ This study calculates realistic target treatmentcosts. Assessing the impact of target costs oncost-effectiveness, however, was beyond thescope of the present study.

Hill A, et al. BMJ Open 2017;7:e011965. doi:10.1136/bmjopen-2016-011965 1

Open Access Research

on June 5, 2020 by guest. Protected by copyright.

http://bmjopen.bm

j.com/

BM

J Open: first published as 10.1136/bm

jopen-2016-011965 on 20 January 2017. Dow

nloaded from

improvements in clinical efficacy.10 The UK’s NationalInstitute for Health and Care Excellence (NICE) has onnumerous occasions in recent years found new cancermedicines to be cost-ineffective compared with currentstandards of care, often because the significantly highercosts are not matched by an improvement in clinical effi-cacy of the same magnitude. Since 2000, 31% of all tech-nology appraisals conducted by NICE for cancer drugsreceived the verdict ‘not recommended’, double theaverage for all treatments.11 For cancer medications,National Health Service (NHS) England has respondedto accusations of ‘rationing’ by creating the controversialCancer Drugs Fund(CDF).12 The CDF provides fundingfor drugs that have not received approval from NICE.Recent analyses of the costs of production for hepatitis

B and C medicines have prompted informed debate onthe optimal provision of treatments and services within aconstrained budget.13 14 This study aims to providesimilar analyses for clinical indications for novel cancermedicines that have been deemed cost-ineffective. Wehave analysed the potential impact of generic import-ation for four drugs, three of which (bortezomib, dasati-nib and everolimus) have been deemed cost-ineffectiveby NICE, and are currently included on the CDF list.15

METHODSCalculation of production costData on active pharmaceutical ingredients (API)exported from India were extracted from an online data-base for 2014 and early 2015.16 Given that prices of APIdecrease with continued market competition, we usedthe lowest per-kilogram API price in this timeframe inour calculations to estimate sustainable generic prices inthe near future.Per-kilogram API prices were input into an algorithm

previously used in analyses of drugs for hepatitis B, C,and oncology drugs.13 14

An example of our calculation algorithm for dasatinibis given in figure 1. The standard dose of dasatinib is100 mg once daily. Thus, the yearly requirement of APIis 36.5 g per patient. The lowest price for dasatinib APIexported from India in 2014 was £1841.14/kg. Theamount of API required to produce one 100 mg tabletwould thus cost £0.18. The total weight of the tablet wasassumed to be five times the weight of the API alone,and excipient prices were calculated by conservativelyassuming that the total non-API mass of the tablet wascomposed of the most expensive excipient. The costs ofexcipients (£0.006 in the case of dasatinib, based onexport data) and tableting (a conservative estimate of£0.026 per tablet) were added to the per-pill cost of theAPI. The resulting per-pill cost of production was multi-plied by 28 to give the monthly cost of production(£6.06/month). Shipping costs and duties at £0.23 permonth, assuming packaging in monthly quantities, wereadded giving a total monthly cost of £6.29. Theseassumptions are based on confidential contact with

generic producers, and would reflect a relatively ineffi-cient manufacturing process. Finally, a 50% mark-up wasadded, to include a profit margin that would incentivisemarket entry and competition between generic manu-facturers, giving a final estimated generic price of £9.43/month, or £122.95 per patient per year.

Patent coverage and global pricesUS basic (substance) patent expiry dates were gatheredfrom the Food and Drug Administration Orange Book.17

Prices for the chosen drugs were identified in elevencountries, using national databases and online price com-parison tools (see online supplementary appendix A).In all cases, the lowest available price per pill was used

Figure 1 Cost estimation flow chart for dasatinib.

2 Hill A, et al. BMJ Open 2017;7:e011965. doi:10.1136/bmjopen-2016-011965

Open Access

on June 5, 2020 by guest. Protected by copyright.

http://bmjopen.bm

j.com/

BM

J Open: first published as 10.1136/bm

jopen-2016-011965 on 20 January 2017. Dow

nloaded from

for comparison. In cases where national pricinginformation was lacking, the corresponding bar is absent(figure 2).

Incidence of cancers and volume demand estimationUsing published figures of the epidemiology of cancersfor which the chosen medicines are indicated, we esti-mated the annual volume of demand in terms of tonnesof API that would be required to treat all incident cases.We estimated the incidence of all cancers for whichthe four chosen drugs are indicated, including multiplemyeloma, chronic myeloid leukaemia, acute lympho-blastic leukaemia, renal cell carcinoma and non-smallcell lung cancer. The potential number of people newlyeligible for treatment with each drug, per year, wasmultiplied by the annual requirement of API in gramsper patient to give annual volume demand.Incidence data for International Classification of

Diseases, tenth edition (ICD-10) categories wereobtained from GLOBOCAN 2012,2 and the incidence ofspecific cancer subtypes was estimated by combiningthese figures with published data from studies on theproportion of cases of the cancer subtype within theICD-10 group. Estimates for the UK were developedusing incidence data from the Cancer Research UK data-base. Taking chronic myeloid leukaemia as an example,it comprises 12.3% of the ICD-10 category ‘leukaemia’.18

For breast cancer, data were only available for women.19

The proportion of incident cases of cancer that wouldbe eligible for treatment with each drug was calculatedby using data on the prevalence of eligibility criteriasuch as the proportion with metastatic disease at presen-tation, or the proportion that are Philadelphiachromosome-positive.

As therapies for clear cell advanced/metastatic renalcarcinoma are not curative, our analysis has assumedthat all patients eligible for first-line treatment will pro-gress and become eligible for second-line treatment witheverolimus.20 For non-clear cell advanced/metastaticrenal cell carcinoma, a consensus on which medicine isfirst-line has not yet emerged, with more than one medi-cine recommended as possible first-line agents.Dasatinib has been recommended as first-line treatmentfor Philadelphia chromosome-positive chronic myeloidleukaemia and Philadelphia chromosome-positive acutelymphoblastic leukaemia.21 22 For the purposes of thisanalysis, all patients for whom everolimus and dasatinibare recommended as one of the possible first-line orsecond-line agents have been included in the eligiblepopulation; our estimates of numbers newly eligible fortreatment with these drugs per year overlap, and wouldbe affected by future changes in treatment guidelines.Our estimates assumed full access to all interventions

indicated before use of drugs, including surgery, radio-therapy and chemotherapy. We do not include measuresof access in our assumptions; where patients do not haveaccess to these interventions, drugs may provide the bestavailable treatment due to low cost, potentially increas-ing the eligible population. In addition, data from high-income countries (HICs) for the proportion of casesthat are advanced/metastatic at presentation are likelyto underestimate the proportion in countries withreduced access to healthcare services and health infor-mation. Finally, our estimates use incidence data, thusgiving the number newly eligible per year. The pointprevalence of eligible people would by definition begreater.

RESULTSCalculated target pricesChemical structures are shown in figures 3 and 4, withreferences for these in online supplementary appendixB. API export data sufficient to allow calculation ofgeneric price estimates were only available for dasatiniband gefitinib (table 1). For bortezomib and everolimus,the lowest priced product globally was used for compari-sons with UK prices.

BortezomibThe recommended dose for bortezomib is 1.3 mg/m2

for a body surface area of 1.8 m2, taken twice a week fortwo consecutive weeks, followed by a resting week, in a3-week cycle. This is equivalent to a per-patient yearlyAPI requirement of 159 mg.The lowest available generic price was for an Indian

product: £199.92 per 3.5 mg phial (figure 5).

DasatinibThe recommended dose for dasatinib is 100 mg takenonce daily, equivalent to a per-patient yearly API require-ment of 36.5 g.

Figure 2 Lowest prices of dasatinib from selected countries.

Hill A, et al. BMJ Open 2017;7:e011965. doi:10.1136/bmjopen-2016-011965 3

Open Access

on June 5, 2020 by guest. Protected by copyright.

http://bmjopen.bm

j.com/

BM

J Open: first published as 10.1136/bm

jopen-2016-011965 on 20 January 2017. Dow

nloaded from

17.5 kg of dasatinib API were exported from India in2014–2015, with the largest volume shipment priced at£1841.14/kg. The most expensive excipient in dasatinibis hypromellose, costing £15.60/kg.The estimated price for dasatinib, assuming a dose of

100 mg daily, was £122.95 GBP per year, or £9.43 GBPper month. The lowest available price was from the ori-ginator company in Brazil, costing £769.03 per month(figure 2).

EverolimusThe recommended dose for everolimus is 10 mg daily,equivalent to a per-patient yearly API requirement of3.7 g. The lowest available generic price globally was£688.96 per month, assuming off-label use, and £851.65on-label, for Indian products (figure 6).

GefitinibThe recommended dose for gefitinib is 250 mg oncedaily, equivalent to a per-patient yearly API requirementof 91.3 g. 416.8 kg of gefitinib API were exported fromIndia in 2014–2015, with the largest single shipmentpriced at £802.56/kg. The most expensive excipient ingefitinib is povidone, costing £9.39/kg.The estimated price, assuming a daily dose of 250 mg,

was £133.73 GBP per year, or £10.26 GBP per month.

The lowest available generic price was £90.49 per month(figure 7).

Patent expiryPatent expiration dates for all drugs are shown intable 2. With the exception of bortezomib, for which thepatent for one particular formulation of the drugexpired in 2014, all drugs are currently under patentprotection. Three of the drugs have multiple active

Table 1 Assumptions and calculations of target prices

Medicine Dasatinib Gefitinib

Daily dose 100 mg 250 mg

Tablets per month 28 28

API price per kilogram £1841.14 £802.56

API cost per tablet £0.18 £0.20

Add cost of excipients £0.19 £0.21

Add cost of tableting £0.22 £0.24

Cost per month £6.06 £6.61

Add cost of bottle, packaging,

shipping, duties

£6.29 £6.84

Add 50% mark-up £9.43 £10.26

Target price per year £122.95 £133.73

The prices of excipients used for each TKI are given in text, butnot shown in table. Bold values are the final target price.EGFR, Epidermal growth factor receptor; TKI, tyrosine kinaseinhibitor.

Figure 3 Chemical structures and formulas for bortezomib

and dasatinib.

Figure 4 Chemical structures and formulas for everolimus

and gefitinib.

4 Hill A, et al. BMJ Open 2017;7:e011965. doi:10.1136/bmjopen-2016-011965

Open Access

on June 5, 2020 by guest. Protected by copyright.

http://bmjopen.bm

j.com/

BM

J Open: first published as 10.1136/bm

jopen-2016-011965 on 20 January 2017. Dow

nloaded from

patents, resulting in a range of expiration dates. Patentexpiry dates were bortezomib 2014–22, dasatinib 2020–26, everolimus 2019–25 and gefitinib 2017.

Global and UK demandIncidence data and assumptions used to calculate eli-gible population estimates are presented in table 3 for

the global population, and in table 4 for the UK popula-tion. References used are given in online supplementaryappendix C.

DISCUSSIONSignificant price reductions can be achieved for numer-ous new cancer medicines, making new treatments avail-able for an estimated 16 611 people in the UK eachyear, for those who live in England, these treatments arenot currently funded by NHS England.Generic production could allow the UK price of dasa-

tinib to decrease by 99.6%, and the UK price of gefitinibto decrease by 99.5%. Importation of Indian genericswould represent a UK price decrease of 74% for borte-zomib and 71% for everolimus. No generic versions ofdasatinib were identified in the countries surveyed.Generic versions of bortezomib were found in India andRussia. Generic everolimus was found in India. Genericgefitinib was found to be available only in India, for £90per month. While this price is significantly below thanthat in other countries (figure 7), it is ninefold the esti-mated generic price of £10 per month. The currentgeneric price of gefitinib in India is roughly equal, peryear, to the median per annum income. It is thereforelikely that the mark-ups set by the generic companiescurrently producing gefitinib are set with marketing to awealthy subset of the Indian population in mind. A lowvolume of demand for gefitinib in India, due to, for

Figure 5 Lowest prices of

bortezomib selected countries.

Figure 6 Lowest prices of everolimus from selected

countries.

Hill A, et al. BMJ Open 2017;7:e011965. doi:10.1136/bmjopen-2016-011965 5

Open Access

on June 5, 2020 by guest. Protected by copyright.

http://bmjopen.bm

j.com/

BM

J Open: first published as 10.1136/bm

jopen-2016-011965 on 20 January 2017. Dow

nloaded from

example, limited state cancer treatment programmes,may also be a contributing factor for the relatively highprice.We estimate that globally, there are 769 736 newly diag-

nosed patients with cancer every year that could betreated with one of these four drugs. Providing thesedrugs to all eligible patients, at target prices, would costan estimated £2.9 billion.The target prices presented in this paper are based on

real-world export and pricing data, calculated using aconservative algorithm that assumes a relatively ineffi-cient manufacturing process and includes shipping andtableting costs, as well as a significant profit margin.Our predictions assume market sizes of a volume suffi-

cient to attract generic producers. For cancer drugs withsmaller patient populations, reductions may be harder

to achieve. Allowing for sufficient demand, and a per-missive legal environment, our findings demonstraterealistic future prices for novel cancer drugs. The pricereductions seen in HIV drugs over the past two decadesshow the dramatic effects of robust generic competitionon access to medicines.23 While our estimates focus onchemically derived medicines, biologics represent agrowing proportion of new cancer medications.24 Thecomplex molecular structures of biologics present regu-latory and manufacturing challenges to the productionof low-cost off-patent biosimilars meaning that, so far,only price reductions of between 10% and 35% havebeen achieved.25 While it may not be possible to achievethe same level of reductions as seen in generics, it islikely that, as manufacturing and regulatory processesmature, and clinicians and patients become more

Figure 7 Lowest prices of

gefitinib from selected countries.

Table 2 Current and target prices

Drug Indication Patent expiry

Current UK drug price

per month (UK)*

Target price

per month

Bortezomib39 1st line MM 2014–22 £762.38 £199.92

Dasatinib40 1st line CML 2020–26 £2504.96 £9.43

Dasatinib41 2nd line CML 2020–26 £2504.96 £9.43

Everolimus42 2nd line RCC 2019–25 £2970.00† £851.65

Everolimus43 Breast CA 2019–25 £2970.00 £851.65

Gefitinib44 1st line NSC lung CA 2017 £2167.71† £10.26

References for patent expiry dates in online supplementary appendix A.*Monthly costs calculated using price from latest version of BNF Online.45

†A Patient Access Scheme (PAS) is in place for this drug. The PAS was not included in our calculations.CA, cancer; CML, chronic myeloid leukaemia; MM, multiple myeloma; NSC, non-small cell; RCC, renal cell carcinoma.

6 Hill A, et al. BMJ Open 2017;7:e011965. doi:10.1136/bmjopen-2016-011965

Open Access

on June 5, 2020 by guest. Protected by copyright.

http://bmjopen.bm

j.com/

BM

J Open: first published as 10.1136/bm

jopen-2016-011965 on 20 January 2017. Dow

nloaded from

Table 3 Global incidence of indicated cancers and estimates of total numbers eligible for treatment with selected medicine

Medicine

ICD-10 category

and incidence

Indication of TKI,

and percentage of

relevant ICD-10

group

Eligibility in terms of

pathology, and percentage

of incident cases with this

subtype

Eligibility in terms of stage of

disease, a percentage of

incident cases at this stage

Total number

newly eligible

for indication,

per year

Total number

eligible

for drug,

per year

Total API

requirement

per year

Bortezomib Multiple myeloma,

114 251

– – Relapsed, received at least 1

prior therapy and who have

already undergone or are

unsuitable for haematopoietic

stem cell transplantation, 25.5%

29 134 143 385 2.6 kg

Multiple myeloma,

114 251

– – Patients for whom high-dose

chemotherapy with stem cell

transplantation is considered

inappropriate, 86.4%

98 713

Multiple myeloma,

114 251

– – Patients for whom high-dose

chemotherapy with stem cell

transplantation is considered

appropriate, 13.6%

15 538

Dasatinib Leukaemia,

351 965

Chronic myeloid

leukaemia, 12.30%

Philadelphia

chromosome-positive, 87.5%

Chronic phase, 90% 34 092 52 280 1.8 tonnes

Leukaemia,

351 965

Chronic myeloid

leukaemia, 12.30%

Philadelphia

chromosome-positive, 87.5%

Intolerant or resistant to

imatinib, 40%

15 152

Leukaemia,

351 965

Acute lymphoblastic

leukaemia, 11.50%

Philadelphia

chromosome-positive, 25%

Refractory to imatinib, 30% 3036

Everolimus Kidney, 337 860 Renal cell

carcinoma, 85%

Clear cell renal cell carcinoma,

77.5%

Advanced/metastatic, 71.5% 159 134 282 678 1.0 tonnes

Kidney, 337 860 Renal cell

carcinoma, 85%

Non-clear cell renal cell

carcinoma, 22.5%

Advanced/metastatic, 71.5% 46 200

Breast, 1 671 149 – Advanced/metastatic, 29.5% HER2 negative, post-aromatase

inhibitor, 12.3%

60 638

Gefitinib Trachea,

bronchus and

lung (C33–34),

1 824 701

Non-small cell lung

cancer, 85%

EGFR-positive, 22.5% Advanced/metastatic, 83.5% 291 393 291 393 26.6 tonnes

Advanced pancreatic neuroendocrine and tuberous sclerosis, for which everolimus is an indicated treatment in some cases, has not been included, due to its relative rarity. Bortezomib isindicated in some cases of mantle cell lymphoma. This has not been included, due to lack of available data.Dosages assumed: bortezomib—two cycles of 1.3 mg/m2 twice weekly for 2 weeks for body surface area of 1.73 m2, dasatinib—100 mg daily, everolimus—10 mg daily, gefitinib—250 mg daily.API, active pharmaceutical ingredients; ICD-10, International Classification of Diseases, tenth edition

HillA,etal.BMJOpen

2017;7:e011965.doi:10.1136/bmjopen-2016-011965

7

OpenAccess

on June 5, 2020 by guest. Protected by copyright. http://bmjopen.bmj.com/ BMJ Open: first published as 10.1136/bmjopen-2016-011965 on 20 January 2017. Downloaded from

Table 4 UK incidence of indicated cancers, and estimates of total numbers eligible for treatment with selected medicine

Medicine

Incidence by

ICD-10 category

Indication of

medicine, and

proportion of

relevant ICD-10

group

Eligibility in terms of pathology,

and percentage of incident cases

with this subtype

Eligibility in terms of stage of disease,

a percentage of incident cases at this

stage

Total number

eligible for

indication,

per year

Total number

eligible for

medicine,

per year

Bortezomib Multiple myeloma,

4792

– – Relapsed, received at least 1 prior therapy

and who have already undergone or are

unsuitable for haematopoietic stem cell

transplantation, 25.5%

1222 6014

Multiple myeloma,

4792

– – Patients for whom high-dose

chemotherapy with stem cell

transplantation is considered inappropriate,

86.4%

4140

Multiple myeloma,

4792

– – Patients for whom high-dose

chemotherapy with stem cell

transplantation is considered appropriate,

13.6%

652

Dasatinib Chronic myeloid

leukaemia, 675

– Philadelphia chromosome-positive,

87.5%

Chronic phase, 90% 532 817

Chronic myeloid

leukaemia, 675

– Philadelphia chromosome-positive,

87.5%

Intolerant or resistant to imatinib, 40% 236

Acute

lymphoblastic

leukaemia, 654

– Philadelphia chromosome-positive,

25%

Refractory to imatinib, 30% 49

Everolimus Kidney, 10 144 Renal cell

carcinoma, 85%

Clear cell renal cell carcinoma,

77.5%

Advanced/metastatic, 71.5% 6165 9780

Kidney, 10 144 Renal cell

carcinoma, 85%

Non-clear cell renal cell carcinoma,

22.5%

Advanced/metastatic, 71.5% 1790

Breast, 50 285 – Advanced/metastatic, 29.5% HER2 negative, postaromatase inhibitor,

12.3%

1825

Gefitinib Lung cancer,

44 488

Non-small cell lung

cancer, 85%

EGFR positive, 22.5% Advanced/metastatic, 83.5% 7104 7104

Advanced pancreatic neuroendocrine and tuberous sclerosis, for which everolimus is an indicated treatment in some cases, has not been included, due to its relative rarity. Bortezomib isindicated in some cases of mantle cell lymphoma. This has not been included due to lack of available data.

8HillA,etal.BM

JOpen

2017;7:e011965.doi:10.1136/bmjopen-2016-011965

OpenAccess

on June 5, 2020 by guest. Protected by copyright. http://bmjopen.bmj.com/ BMJ Open: first published as 10.1136/bmjopen-2016-011965 on 20 January 2017. Downloaded from

familiar with biosimilars, the size of price reductions willincrease in the future.25

Patent expiry dates for the medicines included in thisstudy range from 2014 to 2026. For bortezomib and gefi-tinib, generic competition is likely to be possible in thenext few years, whereas for everolimus and dasatinib,patent protection is likely to prevent the competitionnecessary to reach the target prices. The time to genericmarket entry from patent expiry varies significantlybetween countries. Hudson analysed generic entrybetween 1985 and 1996, finding a range in average timeto entry of between 1.26 and 3.4 years; however, for asample of generics licensed in the EU between 2000 and2007, this ranged from 4 to 7 months, suggesting entry-lag times are decreasing.26 27 There are numerous strat-egies that high, low and middle-income countries canuse to decrease entry-lag. These include supply-side pol-icies such as expedited drug approval processes, anddemand-side policies such as pricing policies.28 29

Several options exist for national governments wishingto facilitate access to medicines by altering the patentstatus. Compulsory License (CL) legislation permits astate to license a patented drug without the patentholder’s consent. Although their use is infrequent, CLsare an effective method of facilitating generic competi-tion, provided for under international agreementssigned by all 161 member countries of the World TradeOrganisation (WTO).30 A CL can only be granted after astate has made meaningful efforts to negotiate a price,unless there is a state of national emergency or ‘extremeurgency’, conditions that the state can determine foritself, in which case the state may proceed directly to aCL. Importantly, the patent holder must still receive rea-sonable remuneration for the CL.31 The WHO has pub-lished guidelines on remuneration of patent holderswhich may help facilitate the pursuit of non-voluntarylicences.32 Relevant domestic legislation may alsoprovide a useful method of negating the barriers posedby patents, because they may provide for different condi-tions to those legislated by the WTO’s Trade RelatedAspects of Intellectual Property rights (TRIPS) agree-ment. In the UK, Crown Use provisions allow the gov-ernment to use or license a patent in the name of thepublic good, and are currently being considered for usewith the monoclonal antibody conjugate, trastuzumabemtansine for refractory breast cancer.33 34 Only dasati-nib, of the drugs included in our study, has been thesubject of CL efforts.35 Even if they are ultimately notrealised, the CL approach may bring price reductions asoriginator companies respond to a change innegotiations.In some cases, voluntary licenses can be agreed

between originator companies and interested thirdparties, facilitating generic production under the termsof license. This approach has most notably been usedwith HIV drugs due to the work of Medicines PatentPool, although it was also used for Gilead Sciences’breakthrough hepatitis C drug, sofosbuvir.36 37

In other cases, patents may be challenged outright.Section 3(d) of the Indian Patent Act allows thirdparties to challenge patent validity, which has in the pastled to the revocation of patents on cancer drugs, andconsequent generic production.38 While it is beyond thescope of this paper to discuss whether these drugs aresuitable candidates for such an approach, it is notablethat dasatinib has been at the centre of a patent disputein India.

CONCLUSIONUsing real-word export data and a conservative manufac-turing model, we calculated realistic target prices forfour cancer drugs. We predict that the resulting pricereductions would have a significant effect on their cost-effectiveness in six clinical indications, making themviable treatment options for more than 750 000 patientsworldwide each year. Some of these clinical indicationsare currently deemed unaffordable by NICE using cost-effectiveness criteria, but if the realistic target price wasavailable, all the drugs may satisfy NICE’s criteria, remov-ing the need for additional funding through initiativessuch as the CDF.Currently, the existing patents on the drugs are the

major barrier to achieving predicted target prices, whichrely on robust generic competition. Numerous strategiesexist for the UK government to pursue in this regard,such as those suggested for the drug trastuzumab emtan-sine. In any case, knowledge of realistic treatment pro-duction costs will be beneficial to price negotiationsacross the world.

Contributors AH designed the study question and methodology. CR draftedthe manuscript. AH and DG provided critique. All authors reviewed andapproved the final manuscript.

Funding This work was supported by an unrestricted research grant fromMetaVirology.

Disclaimer MetaVirology had no editorial control over the final report.

Competing interests None declared.

Provenance and peer review Not commissioned; externally peer reviewed.

Data sharing statement Unpublished export price data for each drug areavailable to interested researchers by emailing the corresponding author. Thedata include shipment size, export destination and cost.

Open Access This is an Open Access article distributed in accordance withthe terms of the Creative Commons Attribution (CC BY 4.0) license, whichpermits others to distribute, remix, adapt and build upon this work, forcommercial use, provided the original work is properly cited. See: http://creativecommons.org/licenses/by/4.0/

REFERENCES1. GBD 2013 Mortality and Causes of Death Collaborators. Global,

regional, and national age-sex specific all-cause and cause-specificmortality for 240 causes of death, 1990–2013: a systematic analysisfor the Global Burden of Disease Study 2013. Lancet Lond Engl2015;385:117–71.

2. Ferlay J, Soerjomataram I, Ervik M, et al. GLOBOCAN 2012 v1.0,cancer incidence and mortality worldwide. IARC CancerBase No.11, 2013. http://globocan.iarc.fr (accessed 1 Feb 2015).

Hill A, et al. BMJ Open 2017;7:e011965. doi:10.1136/bmjopen-2016-011965 9

Open Access

on June 5, 2020 by guest. Protected by copyright.

http://bmjopen.bm

j.com/

BM

J Open: first published as 10.1136/bm

jopen-2016-011965 on 20 January 2017. Dow

nloaded from

3. Beaulieu N, Bloom D, Bloom R, et al. Breakaway: the global burdenof cancer—challenges and opportunities. A report from theEconomist Intelligence Unit, 2009. The Economist. 2009.

4. Hoelder S, Clarke PA, Workman P. Discovery of small moleculecancer drugs: successes, challenges and opportunities. Mol Oncol2012;6:155–76.

5. Knaul FM, Frenk J, Shulman L. Closing the cancer divide:a blueprint to expand access in low and middle income countries.Social Science Research Network, 2011.

6. Savage P, Mahmoud S. Development and economic trends incancer therapeutic drugs: a 5-year update 2010–2014. Br J Cancer2015;112:1037–41.

7. Kelly RJ, Smith TJ. Delivering maximum clinical benefit at anaffordable price: engaging stakeholders in cancer care. LancetOncol 2014;15:e112–18.

8. Lawler M, Le Chevalier T, Murphy MJ, et al. A catalyst for change:the European Cancer Patient’s bill of rights. Oncologist2014;19:217–24.

9. Luengo-Fernandez R, Leal J, Gray A, et al. Economic burden ofcancer across the European Union: a population-based costanalysis. Lancet Oncol 2013;14:1165–74.

10. Kantarjian HM, Fojo T, Mathisen M, et al. Cancer drugs in the UnitedStates: Justum Pretium—The Just Price. J Clin Oncol2013;31:3600–4.

11. NICE statistics | News | NICE. https://www.nice.org.uk/news/nice-statistics (accessed 6 Jun 2016).

12. Jack A. Which way now for the Cancer Drugs Fund? BMJ 2014;349:g5524.

13. Hill A, Khoo S, Fortunak J, et al. Minimum costs for producinghepatitis C direct-acting antivirals for use in large-scale treatmentaccess programs in developing countries. Clin Infect Dis Off PublInfect Dis Soc Am 2014;58:928–36.

14. Hill A, Gotham D, Cooke G, et al. Analysis of minimum target pricesfor production of entecavir to treat hepatitis B in high- andlow-income countries. J Virus Erad 2015;1:103–10.

15. NHS England. National Cancer Drugs Fund List Ver 6.1. 2015.https://www.england.nhs.uk/wp-content/uploads/2016/02/ncdf-list-01-02-16.pdf (accessed 12 Mar 2016).

16. Import Export Data, Export Import shipment data from Customs.http://www.infodriveindia.com/ (accessed 30 Sep 2015).

17. Orange Book: Approved Drug Products with TherapeuticEquivalence Evaluations. https://www.accessdata.fda.gov/scripts/cder/ob/ (accessed 12 Mar 2016).

18. American Cancer Society. Cancer facts & figures 2015. AmericanCancer Society, 2015.

19. Everolimus in combination with exemestane for treating advancedHER2-negative hormone-receptor-positive breast cancer afterendocrine therapy | Guidance and guidelines | NICE. http://www.nice.org.uk/guidance/ta295. (accessed 16 Jul 2015).

20. Ljungberg B, Bensalah K, Canfield S, et al. EAU guidelines on renalcell carcinoma: 2014 update. Eur Urol 2015;67:913–24.

21. Baccarani M, Pileri S, Steegmann JL, et al. Chronic myeloidleukemia: ESMO Clinical Practice Guidelines for diagnosis,treatment and follow-up. Ann Oncol 2012;23(Suppl 7):vii72–7.

22. Alvarnas J, Brown PA, Aoun P, et al. Acute lymphoblastic leukemia.National Comprhensive Cancer Network.

23. Untangling the Web of antiretroviral price reductions: 17th Edition—July 2014. http://www.msfaccess.org/content/untangling-web-antiretroviral-price-reductions-17th-edition-%E2%80%93-july-2014 (accessed 27 Jul 2015).

24. Rugo HS, Linton KM, Cervi P, et al. A clinician’s guide to biosimilarsin oncology. Cancer Treat Rev 2016;46:73–9.

25. Farfan-Portet MI, Gerkens S, Lepage-Nefkens I, et al. Arebiosimilars the next tool to guarantee cost-containment forpharmaceutical expenditures? Eur J Health Econ 2014;15:223–8.

26. Hudson J. Generic take-up in the pharmaceutical market followingpatent expiry: a multi-country study. Int Rev Law Econ 2000;20:205–21.

27. European Commission. Communication from the Commission –

executive summary of the Pharmaceutical Sector Inquiry Report.Reference EU Commission report. 8 July 2009 http://ec.europa.eu/

competition/sectors/pharmaceuticals/inquiry/communication_en.pdf(accessed 17 Mar 2016).

28. Kaplan WA, Ritz LS, Vitello M, et al. Policies to promote use ofgeneric medicines in low and middle income countries: a reviewof published literature, 2000–2010. Health Policy 2012;106:211–24.

29. Kanavos P, Costa-Font J, Seeley E. Competition in off-patent drugmarkets: issues, regulation and evidence. Econ Policy2008;23:500–44.

30. Beall R, Kuhn R. Trends in compulsory licensing of pharmaceuticalssince the Doha declaration: a database analysis. PLoS Med 2012;9:e1001154.

31. WTO | intellectual property (TRIPS). TRIPS and public health:Compulsory licensing of pharmaceuticals and TRIPS. https://www.wto.org/english/tratop_e/trips_e/public_health_faq_e.htm (accessed13 Jun 2016).

32. Love J. Remuneration guidelines for non-voluntary use of a patenton medical technologies. World Health Organization Dept ofTechnical Cooperation for Essential Drugs and Traditional Medicine,2005.

33. The Patents Act 1977—Publications—GOV.UK. https://www.gov.uk/government/publications/the-patents-act-1977 (accessed 30 Sep2015).

34. Boseley S. Health secretary urged to tear up patent on breastcancer drug. The Guardian, Oct 2015. https://www.theguardian.com/society/2015/oct/01/health-secretary-breast-cancer-drug-nhs-kadcyla(accessed 10 Dec 2015).

35. Access to Cancer Treatment: A study of medicine pricing issueswith recommendations for improving access to cancer medication| Oxfam GB. Policy & Practice. http://policy-practice.oxfam.org.uk/publications/access-to-cancer-treatment-a-study-of-medicine-pricing-issues-with-recommendati-344070 (accessed 11 Oct2015).

36. Medicines Patent Pool. Progress and Achievements of theMedicines Patent Pool 2010–2015. 1 Jul 2015. http://www.medicinespatentpool.org/wp-content/uploads/WEB_Progress_Report_2015_EN.pdf (accessed 7 Jul 2016).

37. Gilead Announces Generic Licensing Agreements to IncreaseAccess to Hepatitis C Treatments in Developing Countries | Gilead.http://www.gilead.com/news/press-releases/2014/9/gilead-announces-generic-licensing-agreements-to-increase-access-to-hepatitis-c-treatments-in-developing-countries (accessed 11 Oct2015).

38. ‘t Hoen E. A victory for global public health in the Indian SupremeCourt. J. Public Health Policy 2013;34:370–4.

39. Picot J, Cooper K, Bryant J, et al. The clinical effectiveness andcost-effectiveness of bortezomib and thalidomide incombination regimens with an alkylating agent and acorticosteroid for the first-line treatment of multiple myeloma: asystematic review and economic evaluation. Health TechnolAssess 2011;15:1–204.

40. Pavey T, Hoyle M, Ciani O, et al. Dasatinib, nilotinib andstandard-dose imatinib for the first-line treatment of chronic myeloidleukaemia: systematic reviews and economic analyses. HealthTechnol Assess 2012;16:iii–iv, 1–277.

41. Loveman E, Cooper K, Bryant J, et al. Dasatinib, high-dose imatiniband nilotinib for the treatment of imatinib-resistant chronic myeloidleukaemia: a systematic review and economic evaluation. HealthTechnol Assess 2012;16:iii–xiii, 1–137.

42. Pitt M, Crathorne L, Moxham T, et al. Everolimus for the second-linetreatment of advanced and/or metastatic renal cell carcinoma.PenTAG, Peninsula Medical School, 2009.

43. Fleeman N, Bagust A, Beale S, et al. Everolimus in combination withan aromatase inhibitor for the treatment of breast cancer after priorendocrine therapy: a Single Technology Appraisal. LRiG, TheUniversity of Liverpool, 2013.

44. Brown T, Boland A, Bagust A, et al. Gefitinib for the first-linetreatment of locally advanced or metastatic non-small cell lungcancer (NSCLC): a Single Technology Appraisal. LRiG, TheUniversity of Liverpool, 2009.

45. British National Formulary. BMJ Group & Pharmaceutical Press.

10 Hill A, et al. BMJ Open 2017;7:e011965. doi:10.1136/bmjopen-2016-011965

Open Access

on June 5, 2020 by guest. Protected by copyright.

http://bmjopen.bm

j.com/

BM

J Open: first published as 10.1136/bm

jopen-2016-011965 on 20 January 2017. Dow

nloaded from